opengpu-en

Accueil > News > HiPEAC2012 / OpenGPU workshop

HiPEAC 2012 / OpenGPU Workshop

E-mail Print PDF


HiPEAC2012 and OpenGPU are presenting a workshop - 24th and 25th January 2012

Program of OpenGPU Workshop.

 

The OpenGPU project started in 2010. It combines a set of French companies and French academicals research center around the vision of the lack of toolset, framework and expertise for teeming up the GP-GPU. Thus, the participants of the OpenGPU project focus on bringing tools, benchmarks and applications for the GP-GPU target and beyond, helping code production for new hybrid architectures. Bringing the OpenGPU expertise in the HiPEAC conference will help attendees to get a clear vision of GP-GPU benefits and discover its original solutions. The most relevant cases, tools and benchmarks from the OpenGPU project will be presented. Our objective is to reach out to a broader audience than the sole GPU kernel. To this end, we invite laboratories and companies that have a leading expertise in hybrid computing. Our approach of the conference theme reflect the 2 mains aspects of the GP-GPU: first, its requirements in term of power consumption and efficiency; second, its offer, the exaflopic, new HPC and new library.

On the 24th and 25th of January 2012, we will develop a broad spectrum of the GP-GPU use cases.


Download the PDF version of the agenda

 


 

Agenda

Tuesday 24th

 

10:00
Introduction
Wallix (Jean-Noël de Galzain)
10:30
Making GPGPU easier with OpenCL C++ and a flexible architecture (part1)
AMD (Lee Howes, Ben Gaster)
11:00-11:30 Break

11:30
Making GPGPU easier with OpenCL C++ and a flexible architecture (part2)
AMD (Lee Howes, Ben Gaster)
12:00
Prospective vision for the HPC
HP (Patrick Demichel)
13:00-14:30 Lunch

14:30
optimizing code for GPUs with PGI directives (part of OpenACC)
nVidia (Timothy Lanfear)
15:30
Programming Heterogeneous Many-cores Using Directives
CAPS (François Bodin)
16:00-16:30 Break

16:30
Experiments with Spoc
LIP6 (Mathias Bourgoin)
17:00
Automatic OpenCL generation from sequential C programs with Par4All
HPC Project (Ronan Keryell)
17:30
Plugin Eclipse 4PAR4ALL
Obeo
18:00 End

 

Wednesday 25th

 

10:00
Power and GPU consumption
Supélec (Stéphane Vialle)
10:30
Benchmarks for the HPC GPU
Bull (Mathieu Dubois)
11:00-11:30 Break

11:30
Real Time Random Number Testing on GPU
MassiveRand (Eric Mahé)
12:00
GPGPU atmospheric fluid dynamics : SME's point of view and benchmarks
Numtech (Vivien Clauzon)
12:30-14:30 Lunch

14:30
Domain decomposition method on GPU architecture
ECP (Frédéric Magoulès)
15:15
Metaprof : a large scale clustering application with the CCRT Curie HPC
on metagenomic data

INRA & AS+ (Dany Tello AS+,

Fouad Boumezbeur INRA)

16:00-16:30 Break

16:30
The Moonshot project (ARM servers)
HP (Patrick Demichel)
17:00 End

 

 

 

 

 

OpenGPU FUI collaborative research project, with funding from

Logo_EssonneQuadri150x82 BlocMarqueMiEFI_Nov2010 DGCIS_CMJN_bleu_p 2.HDS_CG_HORIZONTAL_RVB


 




Last Updated ( Monday, 27 February 2012 16:30 )